Fix: #966: golang part for CleanPolicy position

Signed-off-by: Peter Pan <[email protected]>

pkg/operators/pytorch-operator/apis/pytorch/v1/defaults.go

@@ -16,7 +16,6 @@ package v1
 
 import (
 	"strings"
-
 	common "github.com/kubeflow/arena/pkg/operators/tf-operator/apis/common/v1"
 	v1 "k8s.io/api/core/v1"
 	"k8s.io/apimachinery/pkg/runtime"
@@ -87,9 +86,9 @@ func setTypeNameToCamelCase(job *PyTorchJob, typ PyTorchReplicaType) {
 // SetDefaults_PyTorchJob sets any unspecified values to defaults.
 func SetDefaults_PyTorchJob(job *PyTorchJob) {
 	// Set default cleanpod policy to None.
-	if job.Spec.CleanPodPolicy == nil {
+	if job.Spec.RunPolicy.CleanPodPolicy == nil {
 		policy := common.CleanPodPolicyNone
-		job.Spec.CleanPodPolicy = &policy
+		job.Spec.RunPolicy.CleanPodPolicy = &policy
 	}
 
 	// Update the key of PyTorchReplicaSpecs to camel case.

pkg/operators/pytorch-operator/apis/pytorch/v1/types.go

@@ -51,9 +51,9 @@ type PyTorchJobSpec struct {
 	// +optional
 	BackoffLimit *int32 `json:"backoffLimit,omitempty"`
 
-	// Defines the policy for cleaning up pods after the PyTorchJob completes.
+	// Defines the policy for cleaning up pods( under runPolicy) after the PyTorchJob completes.
 	// Defaults to None.
-	CleanPodPolicy *common.CleanPodPolicy `json:"cleanPodPolicy,omitempty"`
+	RunPolicy *common.RunPolicy `json:"runPolicy,omitempty"`
 
 	// Defines the TTL for cleaning up finished PyTorchJobs (temporary
 	// before Kubernetes adds the cleanup controller).

pkg/operators/tf-operator/apis/common/v1/types.go

@@ -141,6 +141,34 @@ const (
 	CleanPodPolicyNone      CleanPodPolicy = "None"
 )
 
+
+// +k8s:deepcopy-gen=true
+// RunPolicy encapsulates various runtime policies of the distributed training
+// job, for example how to clean up resources and how long the job can stay
+// active.
+type RunPolicy struct {
+	// CleanPodPolicy defines the policy to kill pods after the job completes.
+	// Default to Running.
+	CleanPodPolicy *CleanPodPolicy `json:"cleanPodPolicy,omitempty"`
+
+	// TTLSecondsAfterFinished is the TTL to clean up jobs.
+	// It may take extra ReconcilePeriod seconds for the cleanup, since
+	// reconcile gets called periodically.
+	// Default to infinite.
+	TTLSecondsAfterFinished *int32 `json:"ttlSecondsAfterFinished,omitempty"`
+
+	// Specifies the duration in seconds relative to the startTime that the job may be active
+	// before the system tries to terminate it; value must be positive integer.
+	// +optional
+	ActiveDeadlineSeconds *int64 `json:"activeDeadlineSeconds,omitempty"`
+
+	// Optional number of retries before marking this job failed.
+	// +optional
+	BackoffLimit *int32 `json:"backoffLimit,omitempty"`
+
+}
+
+
 // RestartPolicy describes how the replicas should be restarted.
 // Can be one of: Always, OnFailure, Never, or ExitCode.
 // If none of the following policies is specified, the default one

pkg/operators/tf-operator/apis/common/v1beta2/types.go

@@ -136,6 +136,33 @@ const (
 	CleanPodPolicyNone      CleanPodPolicy = "None"
 )
 
+// +k8s:deepcopy-gen=true
+// RunPolicy encapsulates various runtime policies of the distributed training
+// job, for example how to clean up resources and how long the job can stay
+// active.
+type RunPolicy struct {
+	// CleanPodPolicy defines the policy to kill pods after the job completes.
+	// Default to Running.
+	CleanPodPolicy *CleanPodPolicy `json:"cleanPodPolicy,omitempty"`
+
+	// TTLSecondsAfterFinished is the TTL to clean up jobs.
+	// It may take extra ReconcilePeriod seconds for the cleanup, since
+	// reconcile gets called periodically.
+	// Default to infinite.
+	TTLSecondsAfterFinished *int32 `json:"ttlSecondsAfterFinished,omitempty"`
+
+	// Specifies the duration in seconds relative to the startTime that the job may be active
+	// before the system tries to terminate it; value must be positive integer.
+	// +optional
+	ActiveDeadlineSeconds *int64 `json:"activeDeadlineSeconds,omitempty"`
+
+	// Optional number of retries before marking this job failed.
+	// +optional
+	BackoffLimit *int32 `json:"backoffLimit,omitempty"`
+
+}
+
+
 // RestartPolicy describes how the replicas should be restarted.
 // Only one of the following restart policies may be specified.
 // If none of the following policies is specified, the default one

pkg/operators/tf-operator/apis/tensorflow/v1/defaults.go

@@ -91,9 +91,9 @@ func setTypeNameToCamelCase(tfJob *TFJob, typ TFReplicaType) {
 // SetDefaults_TFJob sets any unspecified values to defaults.
 func SetDefaults_TFJob(tfjob *TFJob) {
 	// Set default cleanpod policy to Running.
-	if tfjob.Spec.CleanPodPolicy == nil {
+	if tfjob.Spec.RunPolicy.CleanPodPolicy == nil {
 		running := common.CleanPodPolicyRunning
-		tfjob.Spec.CleanPodPolicy = &running
+		tfjob.Spec.RunPolicy.CleanPodPolicy = &running
 	}
 
 	// Update the key of TFReplicaSpecs to camel case.

pkg/operators/tf-operator/apis/tensorflow/v1/types.go

@@ -53,7 +53,8 @@ type TFJobSpec struct {
 
 	// Defines the policy for cleaning up pods after the TFJob completes.
 	// Defaults to Running.
-	CleanPodPolicy *common.CleanPodPolicy `json:"cleanPodPolicy,omitempty"`
+	RunPolicy *common.RunPolicy `json:"runPolicy,omitempty"`
+
 
 	// Defines the TTL for cleaning up finished TFJobs (temporary
 	// before kubernetes adds the cleanup controller).

pkg/operators/tf-operator/apis/tensorflow/v1beta2/defaults.go

@@ -91,9 +91,9 @@ func setTypeNameToCamelCase(tfJob *TFJob, typ TFReplicaType) {
 // SetDefaults_TFJob sets any unspecified values to defaults.
 func SetDefaults_TFJob(tfjob *TFJob) {
 	// Set default cleanpod policy to Running.
-	if tfjob.Spec.CleanPodPolicy == nil {
+	if tfjob.Spec.RunPolicy.CleanPodPolicy == nil {
 		running := common.CleanPodPolicyRunning
-		tfjob.Spec.CleanPodPolicy = &running
+		tfjob.Spec.RunPolicy.CleanPodPolicy = &running
 	}
 
 	// Update the key of TFReplicaSpecs to camel case.

pkg/operators/tf-operator/apis/tensorflow/v1beta2/types.go

@@ -55,7 +55,7 @@ type TFJobSpec struct {
 	// CleanPodPolicy defines the policy to kill pods after TFJob is
 	// succeeded.
 	// Default to Running.
-	CleanPodPolicy *common.CleanPodPolicy `json:"cleanPodPolicy,omitempty"`
+	RunPolicy *common.RunPolicy `json:"runPolicy,omitempty"`
 
 	// TTLSecondsAfterFinished is the TTL to clean up tf-jobs (temporary
 	// before kubernetes adds the cleanup controller).